Looking forward to some interesting insight

[Arijit Ghosh]

Hi Jared, Hi William

Almost all the reviews in Amazon (though not many) look very impressive. Eager to find more about this book.

Looking forward to interesting postings starting tomorrow.

Regards,
Arijit

[Saurabh Gour]

This is probably the most important subject and it seems it is gaining importance with the inclusion of more organisations joining the process. The team 'corporate' is realising the real value that can be achieved by using software for automation of processes. But as new entrants come in, they all are not mature in organising and facilitating the effort. Though I am sure they would like to. So the responsibility is shifting more on the software vendors to make sure that they help the client with effective means and ways of optimum achievement. This book promises a good deal in order to achieve this goal.

[Jared Richardson]

Originally posted by Arijit Ghosh:
Hi Jared, Hi William

Almost all the reviews in Amazon (though not many) look very impressive. Eager to find more about this book.

Looking forward to interesting postings starting tomorrow.

Regards,
Arijit

Thanks! I've a blog up with a few posts covering a few questions mailed in by readers, topics that didn't make it into the book, etc. It might be a good place to get a jump on the crowd.


http://www.jaredrichardson.net/blog/

[Arijit Ghosh]

Hi Jared,

Checked up the blog.

Also checked up pragmatic website.

Introduction is very catchy . If the complete book lives up to the facts mentioned in the Introduction, then the book is definitely going to be very interesting.

I am expecting lot of Practical Case Studies on multiple projects covering different daily scenarios. One good point for the book would be if it covers different company sizes as mentioned in the Introduction.

A little information about the kind of tools, techniques and the best practices being used to reach to such conclusions would make it more realistic.

Do we get to see all this ?

Regards,
Arijit

[Jared Richardson]

I am expecting lot of Practical Case Studies on multiple projects covering different daily scenarios. One good point for the book would be if it covers different company sizes as mentioned in the Introduction.

We tried to include stories that were informal. I'm not sure I'd call them case studies but we tried to add examples. If there are specific things in the book that you'd like to hear more about, visit the Ship It! mailing list and ask the question.

A little information about the kind of tools, techniques and the best practices being used to reach to such conclusions would make it more realistic.

Yes. Each chapter talks about the tools or ideas used to solve each problem. We added a section called Danger Signs to warn you when you've gone off course with a practice. There are also sections called "How Do I Get Started" and "Am I Doing This Right".

We did keep the book language neutal but we also strived to make it practical and useful.

[Tony William]

Like the author's mind set on "stories that were informal". Formal case studies will usually make thing worse.

I believe every member in every project team have their own stories.... as long as the book makes the point clear, throwing out new / interesting idea, the reader should be able to map it to their own experience / stories.

[Arijit Ghosh]

Hi Jared,

I was reading "tech_tips" extract from the book that is available online on your website.

I could dierctly relate it to our daily develoment environment. We maintain central versioning repository - and is the sole and heart of our effective business strategy.

Daily commits and updates ensures that the code is compatible to that of the rest of the team and modifications are correctly reflected in releases.

Versioning aspect helps in back tracking when required and also a detailed Server and Local history helps in retrieving almost any version of the file during its dvelopment process.

One drawback - backups. CVS being the central repository is a huge database and taking periodic backups are time consuming and most often redundant (but necessary). Minor changes to maybe one or 2 files, results in abcking up the whole project.

Do you have any suggestion or workarounds for such a scenario ?

Regards,
Arijit

[Jared Richardson]

I'd first ask about how you are backing up CVS. If you copy to another disk (over the network? external w/firewire?), you can schedule the backup to run in the middle of the night. Unless your database is measured in tens of gigabytes (or larger), I think this would handle the performance concerns.

Also, there tools available to keep two CVS repositories synchronized. I'd have to search to find them, but if memory serves, you can install the sync process on your primary CVS server and have it mirror all the commits to the second CVS server. The second CVS server is your backup. If you need to take down the CVS server to do the backup, you should only have to take the secondary CVS server down, leaving your primary available for the developers.

You could also do something as simple as install a RAID 1 configuration where all disk writes are written to both disks. The second disk is a mirror of the first. The second disk is your backup.

[Tony William]

Jared,

Are you talking about the tool CVSync (www.cvsync.org)? Is there a build / package that can be downloaded and run on Win32 platform?

[Jared Richardson]

Originally posted by Tony William:
Jared,

Are you talking about the tool CVSync (www.cvsync.org)? Is there a build / package that can be downloaded and run on Win32 platform?

Tools like CVSync, yes, but I didn't have a particular one in mind. As to Win32, for some reason most of the CVS tools seem to be unix only. Given the price of Linux that's never been a real problem for me though.

[Arijit Ghosh]

Hi Jared,

Although the CVS server could definitely be Linux based, but as the developers are mainly using Windows.. so primarily a Windows based effective CVS client is very important.

Anyone has anything in mind ? Does your book harp on the effective process management with regard to centralising data, verisoning them and maintaining an ever-growing repository ?

Regards,
Arijit

[Jared Richardson]

Although the CVS server could definitely be Linux based, but as the developers are mainly using Windows.. so primarily a Windows based effective CVS client is very important.

I agree. Many of today's IDEs have built in CVS clients (like Eclipse does). Two other popular options are TortiseCVS and WinCVS. The last option (that I keep coming back to myself) is command line CVS. I really think people ought to learn with command line CVS so that they understand what the GUIs are doing for them. Otherwise the GUI becomes a Evil Wizard (from The Pragmatic Programmer). That's a GUI that keeps you so insulated from the details that you don't know what's really going on, and when something breaks you are completely lost.

Anyone has anything in mind ? Does your book harp on the effective process management with regard to centralising data, verisoning them and maintaining an ever-growing repository ?

The book focuses on using source code management with a central server and getting it backed up, but it doesn't have management strategies for large repositories.

Ship It! is more focused on getting people using the basic practices (like source code management, continuous integration, test automation, etc) than it does advanced management techniques for the various products.